[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
RE: PROTO Process
Hi,
I think the why is simple - if the normal process works adequately,
the shepherd has very little work to do other than monitor the process
for problems.
If DISCUSSes are documented and communicated on a timely basis to the
authors and chairs, then the shpeherd has no work to do there.
If the authors respond quickly and effectively to the DISCUSSes and
COMMENTs, then the shpeherd has no work to do there.
If the IESG and the authors/chairs reach satsifactory agreements on
requested changes, in a timely manner, then the shepherd has no work
to do.
I would be more concerned if these "issues" did not happen rarely
(where "issues" and "rarely" depends on your interpretation).
As co-chair, with co-chairs acting as shepherds, we recently had five
documents go through the approval process. We did run into issues when
(multiple) DISCUSSes were at odds with the WG consensus. The
chairs/shepherds worked with the ADs and the WG to resolve the
problems and get the documents modified on a timely basis and get them
approved (well, we have one still in process and the shepherd (me) is
being somewhat ineffective). It probably added an extra year of work
addressing the DISCUSSes, but the ADs, the chairs/shepherds, the
authors, and the WG together kept things moving forward. The shepherd
jumped in as needed, but to a large degree, while it was very
frustrating dealing with the amount of changes and constant delays,
the process was a great success because everybody helped keep things
moving.
As author, I have recently had three documents go through the process.
They drew multiple DISCUSSes, some new and some of the
here-we-go-again variety, and we worked them out quickly and, because
the process worked well, we didn't really need shepherd intervention.
I did look to see who the shepherd was at one point because I thought
we might need an arbiter. The IESG, the chair/shepherd, the authors,
and the WG worked together to resolve all the issues on a timely
basis. Again a nice successful process that did not require shepherd
intervention.
As contributor, I recently participated in documents going through the
approval process, and there were some issues raised, and the
(non-chair) shepherd was instrumental in keeping things moving forward
when they might have gotten sidetracked. Another process success.
So I'm thrilled that we do not see more issues that require a shepherd
to resolve. And when we have in my experience, the shepherds have
worked well.
I don't understand why the IESG thinks PROTO has problems. It seems to
me to be working well. I think this is a desirable state of affairs.
David Harrington
dbharrington at comcast.net
ietfdbh at comcast.net
dharrington at huawei.com
> -----Original Message-----
> From: Romascanu, Dan (Dan) [mailto:dromasca at avaya.com]
> Sent: Tuesday, May 12, 2009 8:57 AM
> To: David Harrington; Jari Arkko; Russ Housley
> Cc: wgchairs at ietf.org
> Subject: RE: PROTO Process
>
> The issues raised by David are exactly the added value that I would
> expect from the PROTO shepherds (WG chairs or not) beyond the
> write-up.
> The issue is that this happens so rarely. The question is why.
>
> Dan
>
>
> > -----Original Message-----
> > From: iesg-bounces at ietf.org [mailto:iesg-bounces at ietf.org] On
> > Behalf Of David Harrington
> > Sent: Tuesday, May 12, 2009 3:43 PM
> > To: 'Jari Arkko'; 'Russ Housley'
> > Cc: wgchairs at ietf.org
> > Subject: RE: PROTO Process
> >
> > Hi,
> >
> > Maybe for the IESG, the writeups are the most important and
> > useful aspect. As an author and chair, I think the PROTO
> > process is working just fine.
> >
> > As an author, I expect to deal directly with ADs and
> > reviewers who make comments or DISCUSSes. I think that works
> > just fine most of the time.
> >
> > However, sometimes it does not work that well if the author
> > thinks the change should not be made, and a DISCUSS insists
> > it must. Sometimes, a WG totally overlooks some aspect, such
> > as congestion control or operational issues, and the IESG
> > rightfully thinks the aspect should be addressed. Sometimes
> > the WG does not want to deal with delay that would occur from
> > addressing the issue. Sometimes addressing the aspect can
> > have significant effect on the protocol design. The shepherd
> > can help to determine whether WG pushback is to avoid doing
> > the extra work, or because there is a real technical reason
> > for not doing the change. The shepherd can help to ensure
> > that the WG is kept in the loop on the proposed changes, and
> > not just the IESG and the authors.
> > As an editor, I like knowing the shepherd (and chairs) are
> > there to make sure all the parties have adequate
> > opportunities for input to the process. As a chair, I know it
> > can be easy to be driven to get the milestones completed.
> > Having a shepherd (even if it is one of the
> > chairs) means that somebody has the explicit job of ensuring
> > fairness and transparency during the approval process. I
> > appreciate that the shpeherd is there for that.
> >
> > IESG members typically get involved in the process very late,
> > and do not always understand what debates have already
> > occurred in the WG and were settled only by rough consensus.
> > Those who "lost" the debate can raise the same issues during
> > IETF last call or IESG review to try once again to force
> > their viewpoint through. If the requested change is not
> > consistent with WG rough consensus, then the shepherd can
> > serve as the arbiter. The shepherd can lead the discussion to
> > understand the viewpoints and the history of the two sides,
> > and to help educate the IESG about the history of the issue.
> > Of course the WG chairs also serve that function, and the
> > shepherd should certainly call upon the chairs as needed. As
> > an editor, I appreciate knowing that such arbiters are
> > available to help settle disputes that might arise
> > (again) during IESG reviews.
> >
> > By the time many documents reach the IESG, the chairs and
> > authors are often tired of the document. They just want it
> > done. And they can be frustrated by delays late in the
> > process. The shepherd helps to keep driving the approval
> > process forward, making sure it is clear what and who are in
> > the critical path - whose court is the ball in? What needs to
> > be done to keep things moving. I appreciate knowing that the
> > shepherd is doing this job. I do think this is most helpful
> > when the shepherd is not one of the WG chairs, so they can
> > help to motivate the chairs to get things done on a timely basis.
> >
> > So, personally, as an author and chair, I think the PROTO
> > process is working effectively as designed. I think it may
> > offer more visible benefits to authors and chairs than to
> > IESG members.
> >
> > David Harrington
> > dbharrington at comcast.net
> > ietfdbh at comcast.net
> > dharrington at huawei.com
> >
> >
> > > -----Original Message-----
> > > From: wgchairs-bounces at ietf.org
> > > [mailto:wgchairs-bounces at ietf.org] On Behalf Of Jari Arkko
> > > Sent: Tuesday, May 12, 2009 6:14 AM
> > > To: Russ Housley
> > > Cc: wgchairs at ietf.org
> > > Subject: Re: PROTO Process
> > >
> > > I wanted to expand a little bit on what Russ was saying.
> > >
> > > In my experience, the part of PROTO that is currently
> working is the
> >
> > > write-ups. When it comes to the shepherding of the
> document through
> > > issues raised in last call, directorate, and IESG reviews,
> > the authors
> > > and the responsible/discussing ADs seem to be most active in
> > > responding, explaining, suggesting text changes, etc. There are
> > > notable exceptions
> > > -- you know who you are -- but in many cases the shepherds are
> > > relatively silent. Is this a correct perception of the
> > reality? And if
> > > it is, is it a desirable state of affairs or something that
should
> > be
> > > changed?
> > >
> > > Jari
> > >
> > >
> >
> >
>